Performance Comparison of GeneXpert MTB/RIF, Gene Chip Technology, and Modified Roche Culture Method in Detecting...

Our aim of this study was to observe and analyze the performance of the real-time fluorescence quantitative nucleic acid amplification detection of Mycobacterium tuberculosis/rifampicin resistance (GeneXpert MTB/RIF), gene chip technology, and modified Roche culture method in detecting MTB in sputum submitted for inspection and drug susceptibility. Patients with smear-negative suspected pulmonary TB (n = 120) in our hospital were enrolled in this study using a random number table, and sputum samples submitted for inspection were tested by the GeneXpert MTB/RIF, gene chip technology, and modified Roche culture method. With clinical diagnosis as the gold standard, the performance (mainly sensitivity and specificity) of the above three detection methods in the diagnosis of MTB was compared. Next, the drug susceptibility test (DST) was carried out on sputum samples, tested positive by the three methods. With the solid culture results as the evaluation criteria, the performance of the three detection methods in the diagnosis MTB and DST was compared. When compared with the modified Roche culture method, the GeneXpert MTB/RIF had the highest positive rate and a shorter overall test duration (P < 0.05). In contrast with the gene chip technology, the GeneXpert MTB/RIF exhibited higher sensitivity and negative predictive value (NPV) and lower specificity, accuracy, positive predictive value (PPV), and Kappa value (P < 0.05). According to analysis of the diagnostic performance of the three detection methods, GeneXpert MTB/RIF displayed the highest diagnostic sensitivity, ideal predictive values, and the highest similarity with clinical diagnosis in results (P < 0.05). The detection of susceptibility to isoniazid (INH) and RIF showed that the GeneXpert MTB/RIF and gene chip technology performed ideally in DST of MTB. In comparison with the modified Roche culture method, the GeneXpert MTB/RIF and gene chip technology have more prominent performance in detecting MTB and drug susceptibility. Besides, to further improve the accuracy of clinical diagnosis, various molecular biology detection methods can be combined to avoid delaying of the best time for the diagnosis and treatment of the disease.